Differences

This shows you the differences between two versions of the page.

--- guide:cloud:faq [2020/08/19 17:10]
lphukungoane [FAQ]
+++ guide:cloud:faq [2021/12/09 16:42] (current)
@@ Line 9: / Line 9: @@
     *[[guide:cloud:faq|FAQs]]
-======FAQ======
+======Technical FAQs======
-**Q: Do you monitor my VM’s activity?**
+) ** Do you monitor my VM’s activity?**
-A: We do not monitor user VM activity, but we used in the past a free open-source monitoring tool called Sensu to monitor our own VMs, and we have an Ansible playbook that you could use to deploy the sensu clients and server: https://github.com/ICGC-TCGA-PanCancer/monitoring-bag
+     Answer: We do not monitor user VM activity.
-Sensu can be extended and you can send some of its metrics (cpu, memory, disk, network) to a Graphite server. There are are many other open-source monitoring solutions (Nagios, Cacti, Zabbix, etc) that can be used.
-**Q: I cannot SSH into my VM which was working fine yesterday, do you know what happened?**
-A: Please provide us more details about your VM in order to investigate. In the meantime, please click on the Instance name in the Compute -> Instances tab, and then click on the Console tab to see if there is any error output sent to the console.
+) ** I cannot SSH into my VM which was working fine yesterday, do you know what happened?**
-If you see messages like below it means that the kernel has hung and you will have to hard reboot the VM using the Actions drop-down menu.
+     Answer: Please provide us more details about your VM in order to investigate. In the meantime, please
+     click on the Instance name in the Compute -> Instances tab, and then click on the Console tab to see
+     if there is any error output sent to the console.
+If you see messages like below it means that the kernel has hung and you will have to hard reboot the
+VM using the Actions drop-down menu.
@@ Line 35: / Line 38: @@
    [ 2040.401883] INFO: task apt-check:1286 blocked for more than 120 seconds.
-FAQ - SSH into VM Image
-A possible way to avoid this issue in the future is described in this blog: https://www.blackmoreops.com/2014/09/22/linux-kernel-panic-issue-fix-hung_task_timeout_secs-blocked-120-seconds-problem/
 Basically, add these lines to "/etc/sysctl.conf":
@@ Line 54: / Line 51: @@
 A good reading describing these parameters can also be found here: https://lonesysadmin.net/2013/12/22/better-linux-disk-caching-performance-vm-dirty_ratio/
-**Q: How can I save the changes I made to my VM, so I can use it later?**
-A: You can snapshot your instance which saves its state into a new base image that you can use later. The snapshot causes the instance to be paused temporarily and it's preferable to not have disk activity while doing this, and the snapshot might even fail if the instance has (or had) a lot of data written to it since it was started. Please see the User Guide section for more info on how to take a snapshot.
+) ** How can I save the changes I made to my VM, so I can use it later?**
-**
-Q: What can I do to protect my VM by accidentally being terminated by other colleagues sharing the resources available to our project?**
-A: You can “lock” the instance which will cause it to not accept any state changes, until it is “unlocked”. The operation is available in the “Actions” drop-down of each instance.
+    Answer: You can snapshot your instance which saves its state into a new base image that you can use
+    later. The snapshot causes the instance to be paused temporarily and it's preferable to not have disk
+    activity while doing this, and the snapshot might even fail if the instance has (or had) a lot of data
+    written to it since it was started. Please see the User Guide section for more info on how to take a
+    snapshot.
-**Q: I attached a large volume to my instance (10 TB), but when I try to format it takes a very long time. What can I do to make thQ: Do you monitor my VM’s activity?**
+) ** I attached a large volume to my instance (10 TB), but when I try to format it takes a very long time. What can I do to make this step faster?**
-A: We do not monitor user VM activity, but we used in the past a free open-source monitoring tool called Sensu to monitor our own VMs, and we have an Ansible playbook that you could use to deploy the sensu clients and server: https://github.com/ICGC-TCGA-PanCancer/monitoring-bag
+Answer: We recommend you format it as XFS and use the "-K" option:
-Sensu can be extended and you can send some of its metrics (cpu, memory, disk, network) to a Graphite server. There are are many other open-source monitoring solutions (Nagios, Cacti, Zabbix, etc) that can be used.
+    mkfs.xfs -K /dev/vdb
-**Q: I cannot SSH intoQ: I cannot SSH into my VM which was working fine yesterday, do you know what happened my VM which was working fine yesterday, do you know what happened**?
+In our tests, it took around 3 min to format a 10 TB attached volume.
-A: //Please provide us more details about your VM in order to investigate. In the meantime, please click on the Instance name in the Compute -> Instances tab, and then click on the Console tab to see if there is any error output sent to the console.
-If you see messages like below it means that the kernel has hung and you will have to hard reboot the VM using the Actions drop-down menu.//
+) ** How do I give access to another user to my instance, without giving him access to my SSH private key?**
+Answer:You should never give another user access to your credentials or private key. Please run the following commands inside your instance in order to create a new user and allow him SSH key-based access to your instance. Replace "username" with your colleague's name.
-[ 1920.399095] Not tainted 3.13.0-79-generic #123-Ubuntu
+    sudo adduser username
-[ 1920.399559] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
+    sudo usermod -a -G users username
-[ 1920.400763] INFO: task apt-check:1286 blocked for more than 120 seconds.
+    mkdir /home/username/.ssh/
-[ 1920.401360] Not tainted 3.13.0-79-generic #123-Ubuntu
+    chmod 700 /home/username/.ssh/
-[ 1920.401820] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
+    chown username:username /home/username/.ssh/
-[ 2040.400079] INFO: task rs:main Q:Reg:1061 blocked for more than 120 seconds.
+    Obtain the user’s SSH public key and add it to his home directory: /home/username/.ssh/authorized_keys
-[ 2040.400697] Not tainted 3.13.0-79-generic #123-Ubuntu
+    chown username:username /home/username/.ssh/authorized_keys
-[ 2040.401132] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
-[ 2040.401883] INFO: task apt-check:1286 blocked for more than 120 seconds.
-FAQ - SSH into VM Image
+----
-A possible way to avoid this issue in the future is described in this blog: https://www.blackmoreops.com/2014/09/22/linux-kernel-panic-issue-fix-hung_task_timeout_secs-blocked-120-seconds-problem/
+**For More Technical Assistance**
+If you can’t find what you are looking for, please contact our support team. We are working hard to get back to you within 1-2 business days. Submit the ticket at [[http://helpdesk.chpc.ac.za|Helpdesk]]
+----
+====== Billing FAQs======
-Basically, add these lines to "/etc/sysctl.conf":
-vm.dirty_background_ratio = 5
+) **I powered off my instance but I was still charged for usage!**
-vm.dirty_ratio = 10
-And then apply them:
+     Answer: CHPC charges for resources allocated to the users whether they are actively using them or not.
+             The resources you were approved for are reserved for you therefore  you obligated to pay
+             for them.
-sysctl -p
+) **How often am I billed?**
-A good reading describing these parameters can also be found here: https://lonesysadmin.net/2013/12/22/better-linux-disk-caching-performance-vm-dirty_ratio/
+    Answer: CHPC will send out invoices on the 5th of every month for the previous calendar month usage.
-Q: How can I save the changes I made to my VM, so I can use it later?
-A: You can snapshot your instance which saves its state into a new base image that you can use later. The snapshot causes the instance to be paused temporarily and it's preferable to not have disk activity while doing this, and the snapshot might even fail if the instance has (or had) a lot of data written to it since it was started. Please see the User Guide section for more info on how to take a snapshot.
+) **If I have an unpaid invoice, does that unpaid invoice get rolled in to the next months invoice?**
-Q: What can I do to protect my VM by accidentally being terminated by other colleagues sharing the resources available to our project?
-A: You can “lock” the instance which will cause it to not accept any state changes, until it is “unlocked”. The operation is available in the “Actions” drop-down of each instance.
+    Answer: Yes.
-FAQ - Protect VM Image
+)** Why does my bill fluctuate each month even though I am consuming the same amount of resources each month?**
-Q: I attached a large volume to my instance (10 TB), but when I try to format it takes a very long time. What can I do to make this step faster?
-A: We recommend you format it as XFS and use the "-K" option:
+    Answer: We charge for resources used per hour and the number of hours per month vary throughout the
+            year (28 days, 30 days, 31 days).
-mkfs.xfs -K /dev/vdb
+) **Why am I not receiving an invoice?**
-In our tests, it took around 3 min to format a 10 TB attached volume.
-Q: How do I give access to another user to my instance, without giving him access to my SSH private key?
-A: You should never give another user access to your credentials or private key. Please run the following commands inside your instance in order to create a new user and allow him SSH key-based access to your instance. Replace "username" with your colleague's name.
-    sudo adduser username
-    sudo usermod -a -G users username
-    mkdir /home/username/.ssh/
-    chmod 700 /home/username/.ssh/
-    chown username:username /home/username/.ssh/
-    Obtain the user’s SSH public key and add it to his home directory: /home/username/.ssh/authorized_keys
-    chown username:username /home/username/.ssh/authorized_keys
-is step faster?
-A: We recommend you format it as XFS and use the "-K" option:
-mkfs.xfs -K /dev/vdb
-In our tests, it took around 3 min to format a 10 TB attached volume.
-Q: How do I give access to another user to my instance, without giving him access to my SSH private key?
-A: You should never give another user access to your credentials or private key. Please run the following commands inside your instance in order to create a new user and allow him SSH key-based access to your instance. Replace "username" with your colleague's name.
-    sudo adduser username
-    sudo usermod -a -G users username
-    mkdir /home/username/.ssh/
-    chmod 700 /home/username/.ssh/
-    chown username:username /home/username/.ssh/
-    Obtain the user’s SSH public key and add it to his home directory: /home/username/.ssh/authorized_keys
-    chown username:username /home/username/.ssh/authorized_keys
+    Answer: CHPC sends cloud resource usage invoices to all paying users on the 5th of every month.
+If you would like to terminate your account please email  [[https://users.chpc.ac.za/helpdesk/ |helpdesk@chpc.ac.za]]
+) **I have a billing related question that is not listed here, help?**
+Answer: Email [[https://users.chpc.ac.za/helpdesk/ | helpdesk@chpc.ac.za]] with your question!

wiki.chpc.ac.za

User Tools

Site Tools

Differences

Page Tools